tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Harsimranjit singh Kler <simran...@gmail.com>
Subject Re: modjk apache response questions
Date Thu, 05 May 2011 18:14:22 GMT
Hi
any other input required ?

On Wed, May 4, 2011 at 9:48 AM, Harsimranjit singh Kler <simran4u2@gmail.com
> wrote:

>
>
>  Hi
>>
>
>
>>
>> i have done testing as required using ab- please find details :.
>>
>> Please let me knw if anything other details required.
>>
>> Is this always manadatory that maxthread in tomcat shoud equal to
>> maxclient(apache)?
>>
>> in my case traffic always distributes among 3 tomcats so i assume to give
>> 500 maxthread on each tomcat(also high values for maxthread in tomcat not
>> recomended)
>>
>>
>     Not sure if following worker setting is proper?
>
>>   *http.conf*
>>
>> <IfModule mpm_worker_module>
>>
>> ServerLimit 50
>>
>> StartServers 2
>>
>> MaxClients 1500
>>
>> MinSpareThreads 50
>>
>> MaxSpareThreads 600
>>
>> ThreadsPerChild 50
>>
>> MaxRequestsPerChild 0
>>
>
>
>>   </IfModule>
>>  *Server.xml(for three tomcat)***
>>
>> <Connector port="8008"
>>
>>                enableLookups="false" redirectPort="8443"
>>
>>                 maxThreads="500" minSpareThreads="25"
>>
>>                maxSpareThreads="75" acceptCount="100"
>>                protocol="AJP/1.3" />
>>
>>
>> *TEST*
>>
>> # ./ab -n 150 -c 50 http://10.58.116.20/test/test.html
>>
>> This is ApacheBench, Version 2.3 <$Revision: 655654 $>
>>
>> Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
>>
>> Licensed to The Apache Software Foundation, http://www.apache.org/
>>
>>
>>
>> Benchmarking 10.58.116.20 (be patient).....done
>>
>>
>>
>>
>>
>> Server Software:        Apache/2.2.17
>>
>> Server Hostname:        10.58.116.20
>>
>> Server Port:            80
>>
>>
>>
>> Document Path:          /test/test.html
>>
>> Document Length:        104 bytes
>>
>>
>>
>> Concurrency Level:      50
>>
>> Time taken for tests:   0.369 seconds
>>
>> Complete requests:      150
>>
>> Failed requests:        0
>>
>> Write errors:           0
>>
>> Total transferred:      52500 bytes
>>
>> HTML transferred:       15600 bytes
>>
>> Requests per second:    406.96 [#/sec] (mean)
>>
>> Time per request:       122.863 [ms] (mean)
>>
>> Time per request:       2.457 [ms] (mean, across all concurrent requests)
>>
>> Transfer rate:          139.10 [Kbytes/sec] received
>>
>>
>>
>> Connection Times (ms)
>>
>>               min  mean[+/-sd] median   max
>>
>> Connect:        0    1   0.3      1       1
>>
>> Processing:     1  112 111.2     84     337
>>
>> Waiting:        1  112 111.2     84     337
>>
>> Total:          1  113 111.3     85     338
>>
>>
>>
>> Percentage of the requests served within a certain time (ms)
>>
>>   50%     85
>>
>>   66%    141
>>
>>   75%    185
>>
>>   80%    215
>>
>>   90%    330
>>
>>   95%    335
>>
>>   98%    336
>>
>>   99%    337
>>
>>  100%    338 (longest request)
>>
>>
>>
>>
>>
>>
>>
>> *]# ./ab -n 3000 -c 1000 **http://10.58.116.20/test/test.html***
>>
>> This is ApacheBench, Version 2.3 <$Revision: 655654 $>
>>
>> Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
>>
>> Licensed to The Apache Software Foundation, http://www.apache.org/
>>
>>
>>
>> Benchmarking 10.58.116.20 (be patient)
>>
>> Completed 300 requests
>>
>> Completed 600 requests
>>
>> Completed 900 requests
>>
>> Completed 1200 requests
>>
>> Completed 1500 requests
>>
>> Completed 1800 requests
>>
>> Completed 2100 requests
>>
>> Completed 2400 requests
>>
>> Completed 2700 requests
>>
>> Completed 3000 requests
>>
>> Finished 3000 requests
>>
>>
>>
>>
>>
>> Server Software:        Apache/2.2.17
>>
>> Server Hostname:        10.58.116.20
>>
>> Server Port:            80
>>
>>
>>
>> Document Path:          /test/test.html
>>
>> Document Length:        104 bytes
>>
>>
>>
>> Concurrency Level:      1000
>>
>> Time taken for tests:   7.246 seconds
>>
>> Complete requests:      3000
>>
>> Failed requests:        0
>>
>> Write errors:           0
>>
>> Total transferred:      1050000 bytes
>>
>> HTML transferred:       312000 bytes
>>
>> Requests per second:    414.00 [#/sec] (mean)
>>
>> Time per request:       2415.471 [ms] (mean)
>>
>> Time per request:       2.415 [ms] (mean, across all concurrent requests)
>>
>> Transfer rate:          141.50 [Kbytes/sec] received
>>
>>
>>
>> Connection Times (ms)
>>
>>               min  mean[+/-sd] median   max
>>
>> Connect:        0   31 277.8      0    3001
>>
>> Processing:     1 1384 2297.0    136    7218
>>
>> Waiting:        1 1383 2296.0    136    7217
>>
>> Total:         57 1415 2308.5    140    7246
>>
>>
>>
>> Percentage of the requests served within a certain time (ms)
>>
>>   50%    140
>>
>>   66%    326
>>
>>   75%   1756
>>
>>   80%   3103
>>
>>   90%   6392
>>
>>   95%   6670
>>
>>   98%   7065
>>
>>   99%   7175
>>
>>  100%   7246 (longest request)
>>
>>
>> On Wed, May 4, 2011 at 2:55 AM, André Warnier <aw@ice-sa.com> wrote:
>>
>>> Christopher Schultz wrote:
>>>
>>>> -----BEGIN PGP SIGNED MESSAGE-----
>>>> Hash: SHA1
>>>>
>>>> André,
>>>>
>>>> On 5/2/2011 12:40 PM, André Warnier wrote:
>>>>
>>>>> He has 3 tomcats though.
>>>>> Or did I misread that ?
>>>>>
>>>>
>>>> Oh, well, there still might be a problem. No httpd configuration was
>>>> shown (only workers.properties), so there are still situations where the
>>>> connections can be way-off.
>>>>
>>>> For instance, if session stickiness is used and you get very unlucky,
>>>> it's possible that two Tomcat backends will sit idle while the third
>>>> will be handling 100% of the traffic. If that instance only has 500
>>>> connections available, there will be 1500 active connections
>>>> bearing-down on it. :(
>>>>
>>>
>>> Yes.  Session stickines is used, and this scenario has been mentioned.
>>> I have been trying to get the OP to use the standard Apache "ab" program,
>>> with a simple request to a static page, to make sure that this was not the
>>> case, in conditions that we can relate to.
>>> Unfortunately, he does not seem convinced by the need to do so, so it is
>>> a bit difficult to help further in that case.
>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
>>> For additional commands, e-mail: users-help@tomcat.apache.org
>>>
>>>
>>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message